300 Scott and Brebner . — On the Anatomy 
It is premature to speculate on the physiological significance 
of the structure we have been considering. There is certainly 
a tendency among many Dicotyledons (both climbers and 
others) to find a more protected position for their conducting 
tissues than is provided by the typical orientation of wood 
and bast. To take only two familiar examples : In Bignonia 
all the larger sieve-tubes are contained in the phloem-plates 
which are so well shielded by the wood. The sieve-tubes of 
the external phloem are comparatively insignificant. Simi- 
larly in Tecoma graiidiflor a, and no doubt in some of the other 
species, the medullary phloem is better developed than the 
normal. In this plant the medullary wood is also characterized 
by its large and numerous conducting elements. Whether the 
internal position of the conducting, and especially of the 
proteid-conducting tissues offers any other advantage than 
that of protection must be left an open question for the 
present. 
Although, as we have seen, Herail’s generalization as to the 
uniform character of the tissue produced on the same side of 
the cambium, is not constantly true, yet we now know that in 
the great majority, even of the anomalous cases, the cambium 
forms phloem centripetally and xylem centrifugally. The 
wood of Strycknos was compared by de Bary and van 
Tieghem 1 with that of the Chenopodiaceae, and the com- 
parison still holds good, though the order of development 
in both is different from what they supposed. As is well 
known, the wood of the Chenopodiaceae, Nyctagineae, Phyto- 
laccaceae, and Aizoaceae shows, within the cambium, concen- 
tric rings of collateral vascular bundles, imbedded in secondary 
ground-tissue, which is itself often lignified. In some cases 
successive cambial rings appear ; in others the secondary 
collateral bundles appear to be all formed by one and the 
same ring of cambium. According to de Bary’s account, 
the phloem of these bundles is formed from an initial 
strand, separated off internally from the cambium 2 , so that in 
1 1. c. P . 797. 
2 1. c. p. 592. 
